Graph Framework for Microservices is a platform software stack that bootstraps and accelerates cloud-native microservice development, that is out-of-the-box ready to thrive in the ever challenging world of distributed systems and SaaS.

This is a Root Cause Analysis framework for microservice systems based on Graph-augumented reasoning and agentic collaboration.

CPAD is a unified framework for anomaly detection in microservice systems that integrates log-trace fusion, contrastive learning, and Positive-Unlabeled (PU) learning. By constructing a Log-Augmented Trace Graph (LAG), CPAD combines semantic log signals and trace topology, enabling fine-grained anomaly detection under weak supervision and noisy con

We propose a lightweight framework for anomaly detection in microservices by combining graph embeddings from GCNs with temporal metrics modeled by LSTM-AEs. Validated on the Alibaba CloudOps 2022 dataset, it achieves high precision with low latency and computational overhead.

We propose a lightweight framework for anomaly detection in microservices by combining graph embeddings from GCNs with temporal metrics modeled by LSTM-AEs. Validated on the Alibaba CloudOps 2022 dataset, it achieves high precision with low latency and computational overhead.

This repository contains the source code and implementation of an automated framework for extracting microservices from monolithic software systems. Using FlaskBB as a primary case study, this tool leverages Large Language Models (LLMs) for semantic understanding and Graph Neural Networks (GNNs) to capture structural dependencies.

AppGraph is a language & framework for expressing an entire web application as a graph of interconnected agentic nodes. Nodes can encapsulate behaviors as complex as a full microservice, and leverages generative AI to perform complex, fully agentic workloads without needing to write any code.